Four coal units at Eraring, two at Bayswater, two more at Vales Point. That’s roughly 8,000 megawatts of dispatchable capacity sitting in NSW that the state’s own roadmap assumed would be gone or going by the early 2030s. Some of it already should have been. Origin Energy’s attempt to close Eraring in 2025 got pushed to 2027 after the NSW government leaned on it, and nobody in the market treats that as the last deadline we’ll see slip.
That’s the frame for NSW’s transition. Not whether coal leaves – it will, but whether the transmission and generation meant to replace it arrives on anything like the same clock.
Right now it isn’t.
A grid built around four ageing coal stations #
NSW still gets the bulk of its electricity from four coal-fired power stations: Eraring, Bayswater, Vales Point and Mount Piper. Between them they’ve underwritten the state’s reliability for decades, and all four are well past the midpoint of their working lives. AGL’s Bayswater is pencilled in for closure around 2033, Mount Piper sits with EnergyAustralia and has a notionally longer runway, and Vales Point’s owners have flagged life extension rather than early retirement. Eraring is the live case study. It’s Australia’s largest power station, it was meant to close in August 2025, and a direct deal between Origin and the NSW government pushed that out by two years with a mix of payments and a floor under the plant’s revenue.
I’d argue that deal told you more about the state of the transition than any AEMO forecast. You don’t pay a generator to keep burning coal longer unless you’re genuinely worried about what happens if it stops. NSW Treasury and the energy department weren’t being cautious for the sake of it. They looked at the replacement pipeline and didn’t like what they saw.
The REZ promise versus the REZ paperwork #
NSW’s answer to the coal problem is five Renewable Energy Zones (Central-West Orana, New England, South West, Hunter-Central Coast and Illawarra) designed to bundle wind, solar and storage behind new transmission so the grid doesn’t choke on connection queues the way it has in Victoria and parts of Queensland. On paper it’s a sound structure. The AEMO Integrated System Plan treats REZs as the backbone of the whole NEM transition, not just a NSW idea, and I’ve covered elsewhere what that plan actually commits the market to – see what the AEMO Integrated System Plan actually says for the detail.
Central-West Orana is the furthest along, and it’s still the best evidence of how hard this is proving. I wrote a status check on it a few months back and the short version hasn’t changed much since: the generation projects are mostly ready, the transmission line that’s meant to carry their output isn’t, and the gap between those two facts is measured in years, not months. See Central-West Orana REZ: NSW’s first REZ for the longer version.
New England is arguably worse off. ACEN Australia’s New England Solar project is generating, Stubbo and Valeria are moving, but the REZ transmission upgrade that was supposed to unlock the zone’s full capacity has been delayed by land access negotiations, cost blowouts and a contractor market stretched thin across three states at once. I covered the mechanics of that in New England REZ: where the wind-solar-storage build actually stands, and separately profiled the developer carrying a lot of the generation risk in ACEN Australia: the Philippines-backed builder behind New England Solar.
Why transmission is the actual bottleneck #
Everyone wants to talk about turbines and panels because they’re visible and they photograph well. The real constraint is steel towers and easements.
Building a 500kV line through farming country means landholder negotiations, environmental approval under state planning law, a construction workforce competing with every other REZ and interconnector project in the country, and a supply chain for transformers and high-voltage cable that’s been stretched since before the pandemic eased. None of that moves at the pace a developer’s financial model assumes. NSW’s transmission operator, Transgrid, has been candid in its own regulatory filings about schedule risk on the REZ links, and the AER’s determinations on what Transgrid can recover from consumers for this build have become their own slow-moving negotiation. I’ve written more broadly about who actually wears that cost in Who really pays for the transmission build-out, and it’s worth remembering that every month of delay on a REZ line is a month where the generation sitting behind it earns less, or gets curtailed, or both.
Batteries have partly filled the gap, which is its own story. The Waratah Super Battery near Munmorah was built explicitly as a system-strength shock absorber for the Eraring-adjacent part of the grid, and it’s been doing real work, not storing bulk energy for arbitrage so much as holding the system together when something trips. I covered its first stretch of operation in Waratah Super Battery: the shock absorber finally earning its keep, and it’s a decent example of the kind of asset that can get built fast precisely because it doesn’t need 200 kilometres of new wire to work.
What the Capacity Investment Scheme is actually buying NSW #
The Commonwealth’s Capacity Investment Scheme has become the main lever pulling new firmed capacity into NSW ahead of coal closures, underwriting revenue for batteries and renewables that might otherwise sit on developers’ spreadsheets waiting for better prices. It’s done real work bringing projects to financial close faster than the merchant market alone would have. Whether it’s picking the right projects, or just the projects ready enough to tender, is a separate question I’ve chewed over in Is the Capacity Investment Scheme quietly picking winners?, and NSW’s allocation under the scheme leans heavily on batteries and firmed wind rather than the kind of long-duration storage the state will eventually need once Eraring’s gone for good.
That’s a reasonable near-term bet. Four-hour batteries cover evening peaks well. They don’t cover a cold, still week in July the way a coal station with coal in the yard does, and NSW doesn’t have a Snowy 2.0-scale pumped hydro project of its own in the pipeline the way Queensland has Borumba. Snowy 2.0 itself sits across the border in regulatory terms but counts toward NSW’s effective reliability, and its own construction troubles (tunnelling delays, cost overruns past the original estimate) are reason enough not to lean on it as the single answer. See Pumped hydro versus batteries: what actually firms the grid for how the two technologies actually compare on duration.
The coal closure date nobody fully trusts anymore #
Eraring’s deferral to 2027 was framed as a one-off bridge. I’d treat it as a precedent instead.
Origin got paid to keep the plant running and NSW got two more years of breathing room on reliability. That’s a rational trade for both sides given where the REZ build actually sits today. But it sets an expectation: if the transmission still isn’t ready in 2027, what’s the argument against doing it again? AGL has been more publicly committed to its Bayswater and Liddell-adjacent closure dates than Origin ever was to Eraring’s, and I’ve covered that posture in AGL Energy: coal’s biggest landlord trying to check out early, but commitments on an investor call are not the same as a plant actually switching off.
The broader question of whether Australia is retiring coal faster than the replacement fleet can absorb isn’t unique to NSW – I’ve argued the national case in Is Australia closing coal faster than it can replace it?, but NSW is the sharpest version of it because it carries the largest single coal fleet left in the NEM and the most complex REZ geography, five zones instead of one or two.
What actually needs to happen next #
The fix isn’t mysterious. It’s sequencing and workforce, mostly.
Transgrid needs the New England and Central-West Orana links energised on something closer to their original schedule, not a revised one that’s already been revised once. The state needs a credible long-duration storage answer beyond Snowy 2.0, because batteries alone won’t cover the kind of still, cold week that used to be Bayswater’s job. And NSW’s own Energy and Climate Change Minister will eventually face the same choice Origin forced on the government with Eraring – pay to keep an old plant running a bit longer, or accept a reliability gap while the wires catch up. EnergyCo NSW, the state body running the REZ rollout, has been reasonably transparent about where each zone sits against its own published timeline – worth checking directly rather than taking a developer’s investor update as the final word.
None of this means the transition stalls. It means the gap between coal leaving and renewables-plus-storage arriving stays open longer than the glossy roadmap suggested, and NSW households and industry carry that gap as higher volatility in the interim. Anyone who tells you it’s a clean handover hasn’t been watching the Transgrid filings closely enough. It’s more like a Test match where the follow-on target keeps moving. You can still win from here, but the margin for error is a lot thinner than the opening session suggested.
Watch Eraring’s actual exit date in 2027, and watch whether EnergyCo’s next quarterly update on New England moves the transmission date forward or back. That tells you more about NSW’s real transition pace than any target published in Macquarie Street.
– Marcus Wren, Editor